Towards visual assessment of succulent valley bushveld

Abstract

Visual vegetation assessment was tested for repeatability and accuracy and compared with an objective survey technique for time and cost efficiency. Eighteen operators first inspected a range of sites where ecological status was known. Using these for reference, they then visually inspected and independently estimated the ecological status of 15 other sites. Single operators could not reliably visually assess ecological status. Repeatability was, however, obtained by using the mean from a group of operators’ independently‐derived estimates. In the worst case, fifteen randomly‐chosen operators would be necessary in an assessment team to distinguish 10% differences in ecological status (P≤0.05). On average, however, only eight operators would be necessary for this degree of precision. A number of operators were found to be biased and they should be removed from assessment teams. Conservative relationships between precision and operator team size are provided for users of this technique who may require a precision other than 10%, or who are limited by the number and type of operators at their disposal. For farm planning and possibly even monitoring change on a farm or regional scale, the visual method is a viable and efficient replacement for current objective techniques.